人工智能子宫颈细胞病理图像辅助诊断系统的应用：一项前瞻性诊断试验研究

谢玲玲; 叶栋栋; 何贵; 林仲秋; 周晖

doi:10.19538/j.fk2026020116

PDF(3641 KB)

中国实用妇科与产科杂志 ›› 2026, Vol. 42 ›› Issue (2) : 212-217. DOI: 10.19538/j.fk2026020116

论著

人工智能子宫颈细胞病理图像辅助诊断系统的应用：一项前瞻性诊断试验研究

谢玲玲 ^1a ,
叶栋栋 ^1a ,
何贵 ^1b ,
林仲秋 ^1a ,
周晖 ^1a^,²

作者信息 +

Application of an artificial intelligence-assisted diagnosis system for cervical cell pathology images：a prospective diagnostic test study

XIE Ling-ling ^1a ,
YE Dong-dong ^1a ,
HE Gui ^1b ,
LIN Zhong-qiu ^1a ,
ZHOU Hui ^1a^,²

Author information +

文章历史 +

摘要

目的通过前瞻性诊断试验研究评估人工智能（artificial intelligence，AI）子宫颈细胞病理图像辅助诊断系统用于诊断子宫颈病变的可行性。方法收集2022年4月27日到2023年3月21日期间就诊于中山大学孙逸仙纪念医院妇科患者的子宫颈液基细胞学样本347例，采用子宫颈细胞病理图像辅助诊断系统进行AI独立阅片或AI辅助TBS诊断，其结果与专家组及人工独立阅片组进行比较，结合最终组织病理学结果评价前期AI辅助诊断模型的准确性。结果（1）以专家组TBS诊断为参考标准，AI辅助阅片组与专家组的一致性最高，加权Kappa值达到0.924。（2）以活检或锥切病理结果为参考标准，AI独立阅片组及AI辅助阅片组诊断≥低级别鳞状上皮内病变（LSIL）及≥高级别鳞状上皮内病变（HSIL）的灵敏度均高于人工独立阅片组及专家组（0.614 vs. 0.614 vs. 0.561 vs. 0.579，0.769 vs. 0.590 vs. 0.564 vs. 0.513）。（3）AI诊断系统联合人乳头瘤病毒（HPV）筛查检出≥HSIL病变的准确率为75.3%，高于人工独立阅片组（67.1%）及参考组（74.0%）。结论该诊断系统用于子宫颈TBS分类有较高的准确性及稳定性，对子宫颈病变可能具有较人工更高的预测价值。

Abstract

Objective To assess the feasibility of an artificial intelligence （AI）-assisted diagnosis system for cervical cell pathtology images in diagnosing cervical lesions through a prospective diagnostic text study. Methods A total of 347 liquid-based cytology samples were collected from patients who visited the Department of Gynecology at Sun Yat-sen Memorial Hospital，Sun Yat-sen University between April 27，2022 and March 21，2023. The samples were independently analyzed by the AI system or with AI assistance in ThinPrep Bethesda System （TBS） classification. The results were compared with those obtained from the expert group and the pure researcher-independent reading group. The accuracy of the AI-assisted diagnosis model was evaluated by taking histopathological findings into account. Results Using the TBS diagnosis of the expert group as the reference standard，the highest consistency was observed between the AI-assisted reading group and the expert group，with a weighted Kappa value of 0.924. When biopsy or conization pathological results were used as the reference standard，the sensitivity of the AI-independent reading group and the AI-assisted reading group in diagnosing ≥LSIL and ≥HSIL diseases was higher than that of the pure researcher-independent reading group and the expert group （0.614 vs. 0.614 vs. 0.561 vs. 0.579；0.769 vs. 0.590 vs. 0.564 vs. 0.513）. The accuracy of the AI diagnosis system combined with HPV screening in detecting ≥HSIL lesions was 75.3%，which was higher than that of the pure researcher-independent reading group （67.1%） and the expert group （74.0%）. Conclusion The AI diagnostic system demonstrates high accuracy and stability in cervical TBS classification, and may possess superior predictive value for cervical lesions compared with manual diagnosis.

导出引用

谢玲玲, 叶栋栋, 何贵, 等. 人工智能子宫颈细胞病理图像辅助诊断系统的应用：一项前瞻性诊断试验研究[J]. 中国实用妇科与产科杂志. 2026, 42(2): 212-217 https://doi.org/10.19538/j.fk2026020116

XIE Ling-ling, YE Dong-dong, HE Gui, et al. Application of an artificial intelligence-assisted diagnosis system for cervical cell pathology images：a prospective diagnostic test study[J]. Chinese Journal of Practical Gynecology and Obstetrics. 2026, 42(2): 212-217 https://doi.org/10.19538/j.fk2026020116

中图分类号： R737.33

参考文献

列表( 原文顺序 | 文献年度倒序 | 文中引用次数倒序 ) 可视化分析

[1]	Siegel RL, Miller KD, Wagle NS, et al. Cancer statistics,2023[J]. CA Cancer J Clin, 2023, 73(1):17-48. DOI:10.3322/caac.21763. 本文引用 [1]

[2]	Han B, Zheng R, Zeng H, et al. Cancer incidence and mortality in China,2022[J]. J Natl Cancer Cent, 2024, 4(1):47-53. DOI:10.1016/j.jncc.2024.01.006. 本文引用 [1]

[3]	Zhang M, Zhong Y, Wang L, et al. Cervical cancer screening coverage - China,2018-2019[J]. China CDC Wkly, 2022, 4(48):1077-1082. DOI:10.46234/ccdcw2022.217. 本文引用 [1]

[4]	Jiang Y, Yang M, Wang S, et al. Emerging role of deep learning-based artificial intelligence in tumor pathology[J]. Cancer Commun (Lond), 2020, 40(4):154-166.DOI:10.1002/cac2.12012. 本文引用 [1]

[5]	狄文, 金明珠. 人工智能技术在妇产科应用中的思考与挑战[J]. 中国实用妇科与产科杂志, 2025, 41(1):11-14.DOI:10.19538/j.fk2025010104. 本文引用 [1]

[6]	Holmström O, Linder N, Kaingu H, et al. Point-of-care digital cytology with artificial intelligence for cervical cancer screening in a resource-limited setting[J]. JAMA Netw Open, 2021, 4(3):e211740. DOI:10.1001/jamanetworkopen.2021.1740. 本文引用 [1]

[7]	Bao H, Bi H, Zhang X, et al. Artificial intelligence-assisted cytology for detection of cervical intraepithelial neoplasia or invasive cancer:A multicenter,clinical-based,observational study[J]. Gynecol Oncol, 2020, 159(1):171-178. DOI:10.1016/j.ygyno.2020.07.099. 本文引用 [1]

[8]	Nambu Y, Mariya T, Shinkai S, et al. A screening assistance system for cervical cytology of squamous cell atypia based on a two-step combined CNN algorithm with label smoothing[J]. Cancer Med, 2022, 11(2):520-529. DOI:10.1002/cam4.4460. 本文引用 [1]

[9]	Cao L, Yang J, Rong Z, et al. A novel attention-guided convolutional network for the detection of abnormal cervical cells in cervical cancer screening[J]. Med Image Anal, 2021,73:102197. DOI:10.1016/j.media.2021.102197. 本文引用 [1]

[10]	Cheng S, Liu S, Yu J, et al. Robust whole slide image analysis for cervical cancer screening using deep learning[J]. Nat Commun, 2021, 12(1):5639. DOI:10.1038/s41467-021-25296-x. 本文引用 [1]

[11]	Zhu X, Li X, Ong K, et al. Hybrid AI-assistive diagnostic model permits rapid TBS classification of cervical liquid-based thin-layer cell smears[J]. Nat Commun, 2021, 12(1):3541. DOI:10.1038/s41467-021-23913-3. 本文引用 [1]

[12]	Park KJ. Cervical adenocarcinoma:integration of HPV status,pattern of invasion,morphology and molecular markers into classification[J]. Histopathology, 2020, 76(1):112-127. DOI:10.1111/his.13995. 本文引用 [1]

[13]	Shin MJ, Cho EY, Jang Y, et al. A beneficial effect of simvastatin on DNA damage in 242T allele of the NADPH oxidase p22phox in hypercholesterolemic patients[J]. Clin Chim Acta, 2005, 360(1-2):46-51. DOI:10.1016/j.cccn.2005.04.001. 本文引用 [1]

[14]	Pimenta JM, Galindo C, Jenkins D, et al. Estimate of the global burden of cervical adenocarcinoma and potential impact of prophylactic human papillomavirus vaccination[J]. BMC Cancer, 2013, 13:553. DOI:10.1186/1471-2407-13-553. 本文引用 [1]

[15]

Xie

, Zhang

, Zhao

, et al. Prior cervical cytology and high-risk HPV testing results for 311 patients with invasive cervical adenocarcinoma:a multicenter retrospective study from China's largest independent operator of pathology laboratories[J]. BMC Infect Dis, 2019, 19(1):962. DOI:10.1186/s12879-019-4614-y.

本文引用 [1]

[16]	马德勇, 冯慧, 贺丹, 等. 子宫颈细胞学非典型腺细胞临床价值分析[J]. 中国实用妇科与产科杂志, 2022, 38(3):331-335.DOI:10.19538/j.fk2022030117. 本文引用 [1]

[17]	李梦园, 王丹丹, 张宁宁, 等. 子宫颈癌筛查方法研究进展[J]. 中国实用妇科与产科杂志, 2023, 39(4):474-477.DOI:10.19538/j.fk2023040119. 本文引用 [1]

[18]	冯慧, 赵撼宇, 赵健, 等. 人工智能识别阴道镜下子宫颈红区在子宫颈癌前病变诊断中的价值[J]. 中国实用妇科与产科杂志, 2025, 41(3):357-360.DOI:10.19538/j.fk2025030120. 本文引用 [1]